Professional Documents
Culture Documents
0100 1042 PDF
0100 1042 PDF
Guide to
Overfill
Prevention
2015 EDITION
Introduction
Why invest?
Key elements
Regulatory requirements
Industry standards
Risk assessment
Proof-testing
Equipment selection
References
ISBN 9789198277906
57599 >
9 789198 277906
“The quality of this book makes it the primary educational tool for the
global process and bulk liquid storage industry to reduce the number of
tank overfills”
Phil Myers,
“If multiple layers of protection such as an independent high level alarm or automatic overfill
prevention system had been present, this massive release [Puerto Rico, 2009 ] most likely
would have been prevented.”
Vidisha Parasram,
Investigator at US Chemical Safety and Hazard Investigation Board (CSB)
Legal disclaimer
This information is provided with the knowledge that the publisher and author are offering generic advice
which may not be applicable in every situation. You should therefore ensure you seek advice from an
appropriate professional.
This book does not contain all information available on the subject. This book has not been created to be
specific to any individual’s or organizations’ situation or needs. Every effort has been made to make this
book as accurate as possible. However, there may be typographical and or content errors. This book contains
information that might be dated. While we work to keep the information up-to-date and correct, we make no
representations or warranties of any kind, express or implied, about the completeness, accuracy, reliability,
suitability or availability with respect to the book or the information, products, services, or related graphics
contained in the book or report for any purpose. Any reliance you place on such information is therefore strictly
at your own risk. Therefore, this book should serve only as a general guide and not as the ultimate source of
subject information. In no event will we be liable for any loss or damage including without limitation, indirect
or consequential loss or damage, arising out of or in connection with the use of this information. You hereby
agree to be bound by this disclaimer or you may return this book.
All rights reserved. No part of this book may be reproduced or transmitted in any form or by any means,
electronic or mechanical, including photocopying, recording or by any information storage and retrieval
system, without written permission from the author.
Table of contents
1. Introduction 11
2. Why invest? 15
3. Key elements 29
4. Regulatory requirements 37
5. Industry standards 43
6. Risk assessment 51
9. Proof-testing 69
12. References 105
Abbreviations
1oo1 One out of one
1oo2 One out of two
2oo3 Two out of three
AOPS Automatic overfill prevention system
BPCS Basic process control system
CH Critical high
ESD Emergency shutdown system
FIT Failures in time; number of failures that
can be expected in one billion (109)
device-hours of operation
FMEDA Failure modes, effects and diagnostic
analysis
HFT Hardware fault tolerance
in situ In place; in the context of overfill
prevention this implies that the
equipment (usually the level sensor)
does not need to be unmounted
IPL Independent protection layer
LAHH Level alarm high-high
LOC Levels of concern
MOPS Manual overfill prevention system
MTBF Mean time between failures
MTTF Mean time to fail
MTTR Mean time to repair
MWL Maximum working level
OPS Overfill prevention system
PFD Probability of failure on demand
PFDAVG Average probability of failure on
demand
RRF Risk reduction factor
SIF Safety instrumented function
SIL Safety integrity level
SIS Safety instrumented system
1 - Introduction
1 Introduction
Topic
1.1
1.2
1.3
1.4
Purpose
Background
Scope
Structure
Page
12
12
13
14
11
1 - Introduction
Picture 1.1 and 1.2: The Buncefield tank overfill accident in 2005 resulted in costs of billions of dollars (this accident is further described in chapter 2.4).
12
1 - Introduction
There is no doubt that safety expectations are the past may not conform to current requirements.
increasing. One reason is that legislators are This book describes the latest advancements in
becoming more aware due to accidents, and as a overfill prevention and how to implement future-
result, regulations and permitting are becoming proof solutions.
increasingly stringent with larger consequences. It
is difficult for the industry to maintain compliance
because solutions that were considered acceptable in
1.3 Scope
Although this book is intended for its defined scope, many of the principles are generic and may therefore be
used elsewhere.
Specific industries The primary target of this book is the following industries:
• Petroleum
• Chemical / Petrochemical
• Power
• Food and beverage
• Pharmaceutical
• Metals and mining
• Airports
Spill causes This book focuses on overfilling. There are a number of other possible causes for tank
spills such as leakage or tank rupture due to corrosion, incorrect couplings or simply
that tank openings have been left open during maintenance. The most prominent
problem is however tank overfills.
Tanks and stored The material presented in this book is applicable to most tank types and applications
products containing liquid hazardous substances (e.g. oil and chemicals), but due to the
generic approach it is impossible to cover every possible application and there are
exceptions such as LNG tanks (Liquid Natural Gas) which are not covered by this
book.
Measurement When filling a tank it is important to be aware of all relevant measurement variables
variables such as pressure, temperature and level. The scope of this book is however limited to
aspects relating to level measurement and associated systems.
13
1 - Introduction
1.4 Structure
This guide is structured to provide impartial
information. The structure is based on the IEC 61511
safety life-cycle. The appendix contains vendor
specific information.
14
2 - Why invest?
2 Why invest?
Topic
2.1
2.2
2.3
Risks related to tank overfills
Probability
Consequence
2.3.1
2.3.2
2.3.3
2.3.4
2.3.5
Life and health
Environmental pollution
Property damages
Corporate social responsibility
Public relations
Page
16
16
18
18
18
18
18
18
2.3.6 Industry damage 18
2.3.7 Legal consequences 18
15
2 - Why invest?
2. Why invest? the fact that some countries are less focused on spill
prevention than others, the true number probably
exceeds 100,000 spills of hazardous products per
year globally. All of these spills do not necessarily
This chapter explains why investment in modern arise from a tank overfill, but the data provides an
overfill prevention is good business because it not interesting perspective.
only reduces the statistically high risk of a tank The insurance company Marsh provides an
overfill but also because it has an immediate positive alternative approach focused only on tank overfills,
financial impact. by collecting actual data from the bulk liquid storage
industry. According to their research on atmospheric
storage tanks , one overfill occurs statistically every
Why invest in modern overfill prevention? 3,300 filling operations (Marsh and McLennan
• Protect life & health Companies, 2001). This equals one overfill every 10
years for a group of 10 tanks where each tank is filled
• Protect environment
3 times per month. Using the same assumptions for
• Protect plant assets a group of 100 tanks, the rate of overfill equals one
• Comply with regulations every year.
• Improve public relations
• Corporate social responsibility Historical industry data indicates:
• Increase plant efficiency
• Minimize financial & legal risks One overfill every 3,300 fillings
2.2 Probability
The probability of a tank overfill can be estimated
using historical data. Although individuals and
companies may try to conceal spills, the United
States environmental agency has been able to report
around 14,000 oil spills annually in the United States
alone (United States Environmental Protection
Agency, 2014). Since the US currently consumes
approximately 20% of the world’s oil demand
(Central Intelligence Agency, 2015), this equates
to 70,000 oil spills globally. Considering there are
hazardous substances other than oil, and original
data is conservative due to spill concealment and
16
2 - Why invest?
As an alternative to referencing historical data, the probability of failure of overfill prevention equipment can
be examined.
Basic tank example with mechanical level and independent switch: What is the
probability of a tank overfill? Once every 2400 fillings.
Assumptions Mechanical level switch
Float and tape level measurement
Tank Operations:
• 2 fillings per month
Calculation
During 10 years or 120 months of operation, 2 fillings each month will add up to 240 fillings in total.
Mechanical float and tape level measurement is stated to fail dangerously undetected once every 5
years. This means 2 failures during 240 fillings. Since repair is expected to occur directly upon transfer
completion, the overall probability of filling with a failed float and tape level measurement is 2/240 = 0.8%.
Similarly, the mechanical level switch is expected to fail dangerously undetected only once during the 240
fillings over 10 years. However, with an annual proof-testing, one must expect that each failure remains
unnoticed for an average of 6 months, which translates to 12 fillings assuming 2 fillings per month. Hence,
the probability of filling with a failed mechanical switch is 12/240 = 5%.
Altogether, the probability of filling a tank with float and tape level measurement AND the mechanical
switch is 0.8% x 5% = 0.04%. Alternatively, once every 1/0,04% = 2,400 fillings.
Fact box 2.1: Basic tank example with mechanical level and independent switch: What is the probability of a tank overfill? Once every 2400 fillings.
17
2 - Why invest?
2.3 Consequence For companies to be viable in the long run they need
to be perceived by the public as operating ethically
Potential consequences of a tank overfill are detailed and correctly according to societal values. Fines,
below, along with case examples in section 2.4. additional regulations and inspections, operational
changes, ownership adjustments and ultimately
2.3.1 Life and health closure are all possible results that can occur if the
public’s trust is lost. Implementing modern overfill
A work environment where there is a probability
prevention is one of many required actions to
of severe consequences such as personal injuries
fulfil the public’s expectation on corporate social
or even fatalities must be avoided at all costs. The
responsibility.
slightest rumor about an unsafe work place or part of
a facility will affect reputation, even if an accident has
not occurred. 2.3.5 Public relations
In cases where an accident occurs that involves The news of an accident spreads quickly. Written
injuries or fatalities, in addition to personal suffering, statements, photos and videos are often made
claims for the company responsible can be expected. available to the public. This can influence regulators
to tighten legislation and increase governmental
A case example of fatalities, injuries and evacuation involvement through additional requirements on
is presented in case 7 “Fatalities, injuries and safety and more frequent and thorough inspections.
evacuation”.
2.3.6 Industry damage
2.3.2 Environmental pollution
An accident does not only affect the responsible
Potential environmental pollution from a tank overfill facility, but also the entire industry. The entire
includes many aspects of the local surroundings. industry is at stake when it comes to incidents.
Drinking water, air pollution, wildlife and the
ecosystems are just a few examples. The local There are numerous examples where a single tank
community’s trust is often closely connected to overfill has affected the entire industry, and a specific
environmental aspects. case example is presented in case 5 “Corporate
fines”.
When an accident occurs that results in
environmental pollution, considerable fines for the 2.3.7 Legal consequences
responsible company may be expected. Additionally
the cost of removing or treating contaminated soil or Tank overfills frequently end up in court or with
water (“clean-up”) can be considerable. settlements involving both criminal and civil
charges. Not only may the responsible company
Case examples of spill clean-up and clean-water be accused, but also its staff, and there are cases
contamination are presented in cases 1 “Spill clean where employees, executives and owners have been
up” and 4 “Clean water contamination”. imprisoned. Here are a few examples:
18
2 - Why invest?
discharge of refuse matter. However, these examples are included to show the
• Jaipur, 2009 (Case 7 “Fatalities, injuries and potential consequences of a tank overfill; the result is
evacuation”): 20 people were accused of one similar independently of how the spill occurred.
or more of the following charges: causing
death by negligence; public servant disobeying
law with intent to cause injury to any person;
punishment of criminal conspiracy; and
punishment for attempting to commit
offences punishable with imprisonment for life
or other imprisonment.
Deflagration
In a deflagration, the combustion process of the
burning wave front initiated at the ignition source
propagates through the flammable mixture at
subsonic speeds. The hazard is the flame or flash
fire that at high temperature has the potential
to burn equipment, people, and ignite other
flammable liquid sources, creating the potential for
fire escalation and other safety hazards.
19
2 - Why invest?
Spill clean-up
Western Massachusetts, United States, 2005
Sequence of events
Small facility with a single operator present while
a bulk liquid storage tank was filled through a
pipeline. The operator thought that he would have
time to go to the bar across the street for a quick
beer. Suddenly the bartender points out that diesel
is shooting out from a tank vent. The operator
runs back to the terminal to close a valve in order
to shut down the flow of incoming product. As a
result of this tank overfill, 23,000 gallons of diesel
Picture 2.2: Spill clean up
was released to the secondary containment which
consisted of soil bottom and steel sides. 14,000 gallons of the released product was recovered using
vacuum trucks and 9,000 gallons were lost to the subsurface which contaminated the groundwater. Light
non-aqueous phase liquid was found in 14 wells during 2 weeks. More than 300,000 gallons of liquids were
extracted and reinjected to recover the soil in the vicinity of the tank. Total cost exceeded $350,000.
Root causes
• Failure to adhere to written instructions
• Incorrect manual calculation of flow-rates
• Overfill prevention system existed but was not automatic
Lessons learned
• Although the personnel were qualified, there was a lack of safety culture that made personnel
deviate from instructions
• An automatic overfill prevention system could have prevented this accident
20
2 - Why invest?
Property damages
Buncefield fuel depot, United Kingdom, 2005
Sequence of events
A floating-roof tank overfilled at a tank terminal which
resulted in the release of large quantities of gasoline near
London. A vapor cloud formed which ignited and caused
a massive explosion and a fire that lasted five days.
The terminal was at the time the fifth largest in the
United Kingdom. The terminal supplied both Heathrow
and Gatwick airports with aviation fuel as well as
distribution of motor fuels and gasoline throughout the Picture 2.3: Property damage caused by the accident in
region. Buncefield
Root causes
The primary root cause was that the electromechanical servo level gauge failed intermittently and the
mechanical level switch used in the independent overfill prevention system was inoperable.
The mechanical level switch required a padlock to retain its check lever in a working position. However, the
switch supplier did not communicate this critical point to the installer and maintenance contractor or the
site operator. Because of this lack of understanding, the padlock was not fitted and as a consequence the
mechanical level switch was inoperable.
The electromechanical servo level gauge had stuck 14 times in the three months prior to this major failure.
The root-cause of the ‘sticking’ was never properly investigated or determined. The lack of a proper
‘lessons-learned’ procedure indicates that there was an obvious problem with the overfill management
system.
Consequences
• 40 People injured but no fatalities
• Major property damages including destruction of tanks and nearby office buildings
• Largest fire in Europe since World War II
• Disruption of nearby transportation routes and businesses
• Groundwater pollution
• Settlements exceeding £700 million (approximately $1 billion)
• Civil and criminal charges against the company and individual employees
Lessons learned
The accident received considerable attention from the public and the government. As a result stringent
regulations were created based on a holistic perspective and the functional safety standard IEC 61511. The
government now inspects that these types of facilities have implemented proper management systems,
risk assessments of all tanks and lessons learned procedures.
More specifically, the Buncefield Major Incident Investigation Board* issued a recommendation to install
an independent automatic overfill prevention system conforming to IEC 61511 on all bulk liquid storage
tanks
Source: Marsh, 2007
21
2 - Why invest?
Bankruptcy
Puerto Rico, United States, 2009
Sequence of events
During the off-loading of gasoline from a tanker ship
to the tank farm, a five million gallon above ground
storage tank overfilled into a secondary containment
dike, resulting in the formation of a large vapor
cloud which ignited after reaching an ignition source
in the wastewater treatment area of the facility. In
addition to causing an extensive vapor cloud fire,
Picture 2.4: Puerto Rico accident in 2009
the blast created a pressure wave registering 2.9
on the Richter scale. For more than two days, dark clouds of particulates and smoke polluted the air, and
petroleum products leaked into the soil and navigable waterways in the surrounding area. The smoke
cloud was large enough to be visible by NASA’s Terra satellite. On the days after the explosion, more than
60 agents from both the FBI and the Bureau of Alcohol, Tobacco, Firearms and Explosives were dispatched
to the site.
Root causes
• Malfunctioning automatic tank gauge (float and tape)
• Lack of independent overfill prevention system
• Incorrect manual calculation of flow rate
• Inadequate overfill management system
• Lack of formal procedures for operations
Consequences
• Bankruptcy
• The blast and fire from multiple secondary explosions resulted in significant damage to the
petroleum storage tanks and other equipment on site and in hundreds of homes and businesses up
to 1.25 miles from the site
• Groundwater pollution
• Calls for additional regulation
• Involvement of the United States Department of Homeland Security (which adds complexity to the
industry)
Lessons learned
One of the aspects that the United States Chemical Safety and Hazard Investigation Board (CSB)
emphasizes is the importance of an independent and automatic overfill prevention system. Additionally,
this incident shows the importance of correctly measuring the actual flow-rate into the tank and an
automatic calculation of the estimated completion time of the transfer. This can be achieved by using a
level transmitter and an automatic calculation of the level rate combined with a calculation of the tank’s
volume
Source: U.S. Chemical Safety and Hazard Investigation Board (2015) and Puerto Rico Seismic Network
(2009)
Case 3: Bankruptcy
22
2 - Why invest?
Sequence of events
Approximately 7,500 gallons of chemicals
used to process coal spilled into the Elk
River in West Virginia from an above ground
storage tank at a small tank depot. The
Elk River is a municipal water source that
serves approximately 300,000 people in the
surrounding area.
Picture 2.5: Water inspection
Root causes
• Corroded tank
• Malfunctioning secondary containment
Consequences
• Officials issued an advice to “do-not-use” the drinking water for 5 days
• The company went bankrupt and the facility was razed to the ground
• Criminal charges against 6 individuals associated to the company (owners, managers and
employees) who pleaded guilty
Lessons learned
Local regulators realized the risk associated with tank overfill and have implemented legislation (United
States Environmental Protection Agency’s “Spill Prevention, Control, and Countermeasure Plan”; SPCC
Plan) for above ground storage tanks. The legislation contains requirements for tank and secondary
containment inspections.
23
2 - Why invest?
Corporate fines
Monongahela River, United States, 1988
Sequence of events
A four-million gallon tank catastrophically
failed. The tank was used for the first time
after being reconstructed at a new site.
One million gallons of the diesel oil spilled
into a storm sewer that flowed into the
Monongahela River.
Consequences
• Federal Government issued a fine
Picture 2.6: Refinery next to river
of $2.25 million, the largest for a
petroleum company at the time
• Lawsuits: one for violating the Clean Water Act and another for violating the Federal Refuse Act
• $18 million in cleanup fees and civil lawsuits from those distressed by the experience.
• The potable water supplies for about 1 million people were disrupted. Water shortages were
common after the incident. Wildlife, fish and mussels were harmed or killed
• Over 1200 residents had to evacuate for approximately a week
Lessons learned
Tank overfills do not only concern the owner of the tank, but also the public. Governments may issue
considerable fines which, including other associated costs with a spill, may result in bankruptcy.
Incidents do not only affect the specific company but also the entire industry. For example this incident is
one of the reasons why the industry, through API, garnered a task group to publish the standard API 653
“Above Ground Storage Tanks Inspector Program”.
24
2 - Why invest?
Condemnation of executives
Texas City refinery, United States, 2005
Sequence of events
The incident occurred during the startup of the raffinate
splitter section of the isomerization unit, when the raffinate
splitter tower was overfilled. The excess gas flowed into a
back-up unit, which then also overflowed and sent a geyser
of gasoline into the air. Flammable liquid was released,
vaporized, and ignited, resulting in an explosion and fire. Picture 2.7: Texas City refinery accident
Root causes
• Malfunctioning level transmitters and alarms
• The level transmitters measurement ranges were insufficient
• Lack of safety culture made the operators regularly deviate from written startup procedures
Consequences
• 15 Contract employees were killed
• A total of 180 workers at the refinery were injured, 66 seriously
• Considerable property damages on-site, and off-site windows were shattered in homes and
businesses located, up to three-quarters of a mile (1.2 km) away from the isomerization unit
• The company was charged with criminal violations of federal environmental laws, and has been
named in lawsuits by the victims’ families. In a quarterly report the company revealed that it had
reserved $700 millions for fatality and personal injury claims, although some cases had not yet been
settled
• Government recommended the company appoint an independent panel to investigate the safety
culture and management systems. The investigation was headed by former United states Secretary
of State James Baker III and the resulting report is known as the “Baker Panel report”
• Victims, media and government officials publicly condemned the company for saving money on
safety while making billions of dollars in profits
Lessons learned
Major accidents are not only costly, but also receive considerable public attention. The company’s
reputation will be damaged and even the acceptance of its existence may be questioned. This also affects
the company’s employees and the executives may personally be liable for the accidents and be publicly
condemned.
According to the US Chemical Safety and Hazard Investigation Board the key issues were:
• Safety culture
• Regulatory oversight
• Lack of process safety metrics
• Human factors
More specifically, the accident could probably have been avoided if level transmitters with a measurement
range covering the entire tank had been installed.
Source: Loren, 2005
25
2 - Why invest?
Sequence of events
During a routine transfer of kerosene between
two terminals, a huge leak occurred at a
“Hammer blind valve”. The liquid rapidly
generated vapors which made it impossible
for the shift operators to address the problem.
After about 15 minutes of the leak starting,
there was a massive explosion followed by a
huge fireball covering the entire installation.
The fire which followed the explosion soon Picture 2.8: Oil fire
spread to all other tanks and continued to rage
for 11 days.
Consequences
• 12 people lost their lives due to burns and asphyxia and more than 300 suffered injuries. Many of the
fatalities were company employees
• Half a million people were evacuated from the area
• The Ministry of Petroleum & Natural Gas immediately thereafter appointed a 5-member committee
to investigate the causes of the accident and submit a report within 60 days
• Accusations were raised against 20 employees
• The police arrested 9 senior company officials including its general manager on charges of criminal
negligence eight months after the accident
Lessons learned
Tank spills can result in very serious consequences including fatalities, injuries and evacuation of nearby
communities.
Source: Oil Industry Safety Directorate, 2010
26
2 - Why invest?
Tank leakages
Leaks occur for a number of reasons, for example corrosion,
tank stress or improper welds. As a part of the ongoing safety
trend, the need for leak detection has increased, and many
countries mandate it by law for certain tank types (e.g. above
ground and underground storage tanks).
By using accurate level measurement, abnormal product
movements in the tank can be monitored and thereby used
to detect leaks. The major advantage with using accurate
level measurement for leak detection is that no additional
equipment is required.
27
2 - Why invest?
2.6 Financial returns of modern 2.6.2.3 Hand-gauging and reading of local level sensors
overfill prevention By having a better understanding of what’s in the
This section describes how the usage of a modern tank, fewer manual measurements are required.
overfill prevention system can generate immediate
2.6.2.4 Maintenance
and long-term financial returns.
Modern overfill prevention systems require less
2.6.1 Increased efficiency maintenance.
28
3 - Key elements
3 Key Elements
Topic
3.1
3.2
3.3
3.4
3.5
3.6
The traditional approach to overfill
prevention
The modern approach to overfill
prevention
Requirements
Risk assessment
Process design
30
31
31
31
3.7 Protection layers 32
3.7.1 Basic process control system 33
3.7.2 Safety layer 33
3.7.3 Passive protection and emergency
response layer 35
29
3 - Key elements
RAGAGEP
Corporate Industry
standards standards
Regulations
Figure 3.1: The origin, and priority, of internal and external overfill prevention requirements
30
3 - Key elements
Risk
Tank fillings Protection layers
Operator
Unacceptable risk
Tolerable risk
Figure 3.2: Basic concept of evaluating the assessed risk compared to the tolerable risk
Modern overfill prevention uses this risk 3.6 Overfill Management System
(performance) based approach as opposed to the
traditional prescriptive approach. This ensures Traditionally, tank overfills have been attributed
that the safeguards are neither over nor under to malfunctioning equipment. Although this is
engineered. often a contributing factor, the actual root-cause is
often more complex and involves human behavior.
An action to reduce the risk to a tolerable level must Therefore a critical part of modern overfill prevention
be taken if the risk assessment determines that the is to establish an adequate Overfill Management
risk is unacceptable. Examples of actions that can be System (OMS) that corresponds with how it works in
taken to reduce the risk: the field.
• Inherent process design change An OMS is the framework of processes and
• Changes in the Overfill Management System procedures used to ensure that the organization
(e.g. operational procedures) fulfills all tasks required to achieve the objective of
• Implementing additional protection layers or tank overfill prevention. This includes components
modifying the existing ones such as competent personnel, written procedures,
lessons learned systems and management of change
Risk Assessment is described further in chapter 6 procedures. Although a large task at first, the
“Risk Assessment”. creation of an adequate OMS is not just a necessity
to prevent tank overfills, it will also result in a more
efficient facility. OMS is described in chapter 7
“Overfill Management System”.
31
3 - Key elements
3.7 Protection layers • All IPLs are not alike. The Basic Process Control
System (BPCS) and Safety layer can be used to
Generally, a multitude of independent protection prevent the accident and thereby reduce the
layers (IPLs) are used to minimize the risk of tank probability, whereas the Passive protection
overfills as depicted in figure 3.3, according to the and Emergency response layers mitigate
principle “do not put all your eggs in one basket”. The the accident and thereby minimize the
commonly used IPLs for tank overfill prevention are consequences.
depicted in figure 3.4 below.
• According to IEC 61511-1 Edition 1: “The risk
To reduce risk, an existing IPL can be modified, or reduction factor for a BPCS (which does not
alternatively an additional IPL can be added. The conform to IEC 61511 or IEC 61508) used as
selection process often involves a cost benefit a protection layer shall be below 10”, this will
analysis. Examples of additional parameters (besides be discussed further in chapter 5 “Industry
internal and external requirements) that should be standards”.
taken into consideration are:
Mitigation
Prevention
Operator
shutdown BPCS DCS or tank inventory software
Figure 3.3: Commonly used independent protection layers (IPLs) to minimize the risk of tank overfills
32
3 - Key elements
Manual Overfill Prevention System (MOPS) Basic Process Control System (BPCS)
LT LT
LC
Secondary
containment
Product inlet
Fire brigade
Manual valve
Pump
Figure 3.4: Generic tank example with the commonly used IPLs for tank overfill prevention and mitigation depicted.
Process
Shutdown Basic process control layer
LT SIS
Time In Use
Value Figure 3.7: Automatic overfill prevention systems (AOPSs) usually consist of
a level transmitter (LT), logic and actuator that automatically closes a valve
Figure 3.5: The BPCS layer is in use continuously and is the primary tool to to prevent overfills from occurring. No human intervention is required which
prevent a tank overfill. The other IPLs are only activated upon failure of the usually increases the reliability and shortens the response time.
subsequent IPL
33
3 - Key elements
Outlook box 3.1: Manual replaced with Automatic Outlook box 3.2: Adherence to IEC 61511
Picture 3.2 and 3.3: Two continuous measurements of the same type are used as the level sensors in the BPCS and
the safety layer (left: radar level gauges, right: guided wave radar level transmitters)
Outlook box 3.3: Continuous level measurement for the safety layer
34
3 - Key elements
A critical aspect of overfill prevention is to correctly 3.7.3 Passive protection and emergency
define the “levels of concern” (LOC) which include response layers
Critical High (CH), Level Alarm High High (LAHH or
simply HiHi) and Maximum Working Level (MWL) as In the case of tank overfill protection, the passive
depicted in figure 3.9. protection layer usually consists of a secondary
containment (e.g. dikes or concrete walls) and the
emergency response layer consists of a fire brigade.
These IPLs are merely used for mitigation of tank
overfills, and they are consequently not included
within the scope of this book.
CH
3.8 Commissioning and subsequent
verification
LAHH
MWL
35
3 - Key elements
3.8.2 Proof-testing
The purpose of proof-testing is to verify that
commissioned equipment already in operation
functions correctly. It is a useful tool to reduce the
safety layer’s probability of failure on demand for
infrequently used safety systems. Proof-testing is
further described in chapter 9 “Proof-testing”.
36
4 - Regulatory requirements
4 Regulatory requirements
Topic
4.1
4.1.1
Different types of regulations
4.1.2
4.1.3
4.1.4
4.2
Prescriptive regulation
Performance based regulation
Page
Implications
40
39
39
40
40
40
37
4 - Regulatory requirements
Flammable/
OSHA combustible
liquids
Spill prevention
Clean water act
Facility response
plan
Regulatory EPA
standards
General duty
clause
Clean air act
Risk management
program (RMP)
USCG Federal
ILTA
IPAA
Figure 4.1: Example mapping conducted by the United States Chemical Safety Board in 2015 of external requirements relating to tank overfills for above
ground storage tank terminals in the United States
38
4 - Regulatory requirements
Performance based
regulations with
extensions
Performance
based regulation
Prescriptive regulation
No regulation
Time
1930 1970 2000
Figure 4.2: Developments of safety regulations in the industrialized world
39
4 - Regulatory requirements
40
4 - Regulatory requirements
Inspired by the developments in the UK, the European Union (EU) has taken the lead in regulations for
process safety through the Seveso directive.
41
4 - Regulatory requirements
42
5 - Industry standards
5 Industry standards
Topic
5.1
IEC 61511
5.1.1
5.1.2
5.1.3
5.2
Basic concepts
IEC 61508 Certification
Page
API 2350
45
45
48
49
50
43
5 - Industry standards
API 2350
BPCS
Figure 5.1: Industry standards IEC 61511 and API 2350 - comparison of intended scopes
44
5 - Industry standards
5.1 IEC 61511: Functional safety – 5.1.1.2 Safety Integrity Level (SIL)
Safety instrumented systems for The reliability of a SIF is quantified in “Safety Integrity
the process industry sector Level” (SIL) 0 – 4, which each corresponds to an
IEC 61511 is intended for safeguards used in interval of its capability to reduce risk, as listed in
the process and bulk liquid industries based on table 5.1.
completely, or partially, electrical/electronic/
programmable components. Safety Integrity Risk Reduction Factor (RRF)
Level (SIL)
SIL 4 >10,000 to ≥ 100,000
IEC 61511
SIL 3 >1,000 to ≥ 10,000
Use this standard for: Automatic Overfill
Prevention Systems in SIL 2 >100 to ≥ 1,000
A single electrical/electronic/programmable
safeguard is denoted “Safety Instrumented Function”
(SIF) and consists of a sensor, logic-solver and
actuator as depicted in figure 5.2.
In the context of overfill prevention, this corresponds
to an automatic overfill prevention system (AOPS)
consisting of one or multiple level sensors, a logic-
solver and one or multiple actuators controlling a
corresponding valve.
45
5 - Industry standards
SIF SIS
SIF
Logic
solver
Sensor
Actuator
Figure 5.3: Principal overview safety instrumented function (SIF) an safety instrumented system (SIS)
46
5 - Industry standards
Safety requirement
specifications for the safety
instrumented system
Installation, commissioning
and validation
Modification
Decommissioning
47
5 - Industry standards
5.1.1.5 Equipment selection maintain the product’s quality. It also places rigorous
requirements on the documentation provided with
IEC 61511 prescribes two different options for the the product.
selection of SIS equipment:
Often, the conformance to IEC 61508 is audited
• IEC 61508 compliant equipment by an independent third party. These assessors
• S
elf-qualification of the equipment based on usually issue a compliance report and a certificate.
the prior use clause in IEC 61511 The value of these certificates is dependent on the
The majority of the industry, with the exception of specific assessor. It is therefore important to ensure
extreme weather locations, use the option of IEC that the assessor adheres to the following minimum
61508 compliant equipment where the manufacturer requirements:
provides a standardized certificate, failure-rate data • Accreditation by a recognized third party
and safety manual. Selecting IEC 61508 equipment
generally results in less work when implementing the • Competency within the field of functional
SIS. safety
• Proper engagement in the development
Self-qualification is a comprehensive process for the project
designer of the SIS that relies on the availability of
relevant historic data for the specific equipment.
Another difficulty for the designer is to develop
testing methodologies and assessing their
effectiveness. Certain vendors provide theoretical
failure-rate data that can be taken into consideration
but it is important to be aware that this does not
alleviate the designer of the SIS from any of the self-
qualification requirements.
48
5 - Industry standards
A product developed according to IEC 61508 implies 5.1.3 IEC 61511 applied to modern overfill
that: prevention
• T
he developer has to have a rigorous Modern overfill prevention requires that automatic
documented management system including: overfill prevention systems (AOPSs) are designed
oo Product development process according to the most recent globally accepted
standard which is currently IEC 61511. This standard
oo Manufacturing process provides a solid framework throughout the life-time
oo Documentation system of the AOPS.
oo Management of change process It is important to understand that IEC 61511 is
oo Lessons learned system focused solely on SIS and therefore does not cover
oo Quality system all the elements of modern overfill prevention (see
chapter 3 “Key elements”). For example, it does not
• During the design of the product the following
cover:
must be included:
oo Failure Modes and Effects Analysis (FMEA) • Internal and external requirements such as
regulations and local standards
oo Comprehensive testing including fault
insertion tests • Process design
oo Documentation that provides traceability • Overfill Management System (e.g. lessons
and evidence for all safety requirements learned procedures)
oo Development of proof-testing procedures • Non-safety layers (i.e. Basic process control
system, Passive protection, and Emergency
• Comprehensive user documentation
response layers)
requirements rendered in a:
The performance (risk) based approach in IEC
oo Safety manual
61511 corresponds to the legislative approach
• Involvement of a third party assessor that “Performance based regulation” but does not
requires and issues: cover all the elements of “Performance based with
oo Audits extensions regulation”.
oo Compliance reports IEC 61511 is one of multiple invaluable elements of
oo Certificates modern overfill prevention.
Products developed and independently assessed for
conformance to IEC 61508 is a lengthy and costly
process for the manufacturer. This assessment
however generates several benefits for the user:
• Quality assurance
• Quantified reliability figures and classification
of safety integrity level (SIL) capability
• Proper documentation covering all parts of the
life-cycle
oo Product information and data (e.g.
reliability and product life-time)
oo Procedures (e.g. installation and proof-
testing)
oo Drawings
49
5 - Industry standards
5.2 API 2350: “Overfill Protection for The standard is a mix of prescriptive and performance
Storage Tanks in Petroleum based requirements. It requires a risk assessment to
Facilities” be conducted and evaluated against the tolerable
risk, while still describing the minimum required tank
With the introduction of the 4th edition, which was overfill equipment on the tank.
a major change compared to previous editions, API
2350 became the first globally recognized overfill A common confusion relates to the standard
prevention standard for the bulk liquid storage in the tank categories that are required to be
industry. determined and the associated minimum equipment
requirements. In practice, most modern facilities are
category 3 according to the API 2350 classification
API 2350 and require the usage of an Automatic Tank Gauging
(ATG) system with independent overfill prevention
Use this standard for: Overfill Protection in
system (OPS). Additionally, when the required
Process Industry risk assessment is conducted it is unlikely that the
determined equipment requirements are lower than
;; Bulk Liquid Storage Industry the API 2350 specified minimum requirements.
Note: API 2350 contains generic principles that API 2350 accepts both MOPS and AOPS, but in case
are also applicable to the process industry sector the latter is used, the basic practical requirement is
(although this is not the intended scope) that it shall be designed according to IEC 61511. The
standard does not place any specific requirement
on the AOPS’s SIL. Instead, this is referred to the risk
The purpose of this standard is to provide a holistic assessment.
perspective that is synchronized with (but does
not cover all parts of) the legislative approach
“Performance based with extensions regulation”
seen in figure 4.2 in chapter 4.
API 2350 contents include:
• Overfill Management System
• Risk assessment
• Operations and procedures
• Overfill Prevention System
• Tank Gauging System
Although API 2350 is generically written, the
intended scope is non-pressurized above-ground
storage tanks containing petroleum products as
defined in table 5.2:
III Flash Point equal to or greater than 140°F (60°C) Home heating oil, Yes - Recommended
lubricating oils, motor oil
Table 5.2: Products included in API 2350’s scope
50
6 - Risk assessment
6 Risk assessment
Topic
6.1
Corporate risk management
6.1.1
6.2
6.2.2
6.2.3
6.3
6.3.1
Tolerable risk
Risk analysis
6.2.1
Hazard identification
Hazard and scenario analysis
Risk
53
53
56
56
56
57
57
57
6.3.2 Identify risk reduction options 57
6.3.3 Prioritization 58
6.3.4 Implementation 58
51
6 - Risk assessment
6. Risk assessment
Risk
Tank fillings Protection layers
Operator
Risk analysis
Hazard identification
Likelihood Consequences
Risk Communication
Corporate risk management
Tolerable risk
Prioritization
Figure 6.2: Basic risk assessment model for
overfill prevention
Implementation
52
6 - Risk assessment
Intolerable risk
Tolerable risk
Consequence
Figure 6.3: Simplified example with tolerable risk
53
6 - Risk assessment
Probability
Therefore the British Health and Safety Executive
(HSE) invented the principle of ALARP which is an
abbreviation of “as low as reasonably practicable”.
Reasonably practicable involves weighing a risk Intolerable risk
against the trouble, time and money needed to
control it. The purpose is to enable proportionate
risk reduction measures and the principle has been AL
AR
widely adopted in the process industry and by other P
countries. An overview of the principle is depicted
in figure 6.4 and the specific numbers used in the Tolerable risk
United Kingdom are presented in figure 6.5 along Consequence
with a comparison in figure 6.6. Figure 6.4: The principle of “as low as reasonably practicable” (ALARP)
Fatalities/man-year
2 3 5 6 7 8 9
Intolerable Risk
Australia (NSW)
Tolerable risk
ALARP Hong Kong
ALARP
Netherlands
Tolerable Risk Intolerable risk
United Kingdom
6.1.1.2 Tolerable risk examples environment and financial losses. The risk graphs
may either be quantitative, semi-quantitative or
The theoretical models described above are typically
qualitative as described in figure 6.7 and 6.8.
implemented by corporations as multiple risk
graphs for the selected consequences, e.g. health,
Consequence
Catastrophic STOP
Catastrophic
Unacceptable URGENT ACTION
Undesirable ACTION
Significant
Acceptable MONITOR
Desirable NO ACTION
Moderate
Low
Negligible
Probability
54
>0.1/yr <0.1/yr <10-2/yr <10-3/yr <10-4/yr
Health Asset Environment Company image
Likely Probable Occasional Remote Improbable
Multiple fatalities Extensive International
(<10-5/yr) damage Massive effect impact Stop Stop SIL 3 SIL 2 SIL 1
(>$10M)
Single fatality Major damage
6 - Risk assessment
(<10-4/yr) (<$10M) Major effect National impact Stop SIL 3 SIL 2 SIL 1 SIL 0
Major injury Major damage Localized effect Considerable SIL 3 SIL 2 SIL 1 SIL 0 OK
55
(<10-3/yr) (<$500K) impact
Minor injury Minor damage Minor effect Minor impact SIL 2 SIL 1 SIL 0 OK OK
(<10-2/yr) (<$100K)
Slight injury Slight damage Slight effect Slight impact SIL 1 SIL 0 OK OK OK
(<0.1/yr) (<$10K)
None None None None OK OK OK OK OK
Figure 6.8: Example corporate risk matrix. Worst-case consequence outcome determines the required risk reduction
6 - Risk assessment
56
6 - Risk assessment
• Possibility of a tank overflow resulting in the estimated financial consequences from the
escalation of hazardous events onsite or single event of product overfill into secondary
offsite containment is estimated as $480,000 in clean-up
• Possibility of impact to nearby sensitive costs. As shown in figure 6.10, this estimation is
environmental receptors based on the assumption that 80% of the product
will be contained by the secondary containment
• Physical and chemical properties of product costing $250,000 (80%x 350,000), 5% will overfill
released during overflow the secondary containment costing an additional
• Maximum potential overfill flow rates and $50,000 (5% x $1,000,000) and a final 15% leakage
duration from the secondary containment will drive yet
A simplified example based on an event analysis is another $150,000 (15% x 1,000,000).
provided in figure 6.10. In this fictitious scenario,
100%
15% Secondary containment Clean-up cost $1m
leaks product
Tank overfill
Event #n
Event #1
57
6 - Risk assessment
58
7 - Overfill management systems
7 Overfill management
system
Topic
7.1
7.2
7.3
Why OMS is needed
The basic elements of OMS
Success factors
Page
60
61
61
59
7 - Overfill management systems
Corporate
Management
System
Safety Overfill
Management Management
system System
Figure 7.1: Venn diagram perspective showing how OMS relates to other
corporate management systems
60
7 - Overfill management systems
61
7 - Overfill management systems
62
8 - Overfill prevention system
Topic
8.1
8.2
8.3
8.4
8.5
Manual overfill prevention system
Automatic overfill prevention
system
AOPS vs. MOPS
Hardware fault tolerance
Levels of concern
65
Page
64
66
66
68
63
8 - Overfill prevention system
Manual Overfill Prevention System (MOPS) Basic Process Control System (BPCS)
LT LT
LC
Figure 8.1: MOPS usually consists of a level transmitter (LT) connected to an audiovisual alarm that notifies an operator to take the appropriate
action, e.g. closing a valve. API 2350 classification: category #3.
64
8 - Overfill prevention system
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
LT LT
SIS LC
Figure 8.2: AOPS usually consists of a level transmitter (LT), logic and actuator which automatically closes a valve to prevent overfills from occurring.
The logic may also execute non-safety critical tasks such as shutting down a pump and notifying the operators through audiovisual alerts. API 2350
classification: category #3.
65
8 - Overfill prevention system
AOPS is a safety instrumented function (SIF) and factor of 10 according to IEC 61511. AOPS in
table 8.1 describes when conformance to IEC 61511 conformance with IEC 61511 can offer risk
is a requirement. reduction factors also above 10
• AOPS can considerably shorten response times
Risk Reduction SIL Conformance to IEC compared to MOPS. It is not unusual that a
Factor 61511 MOPS has a 15 minute response time, whereas
<10 0 Recommended an AOPS has below 1 minute
• MOPS requires personnel in the field in
>10 1,2,3,4 Required
potentially unsafe working conditions
Table 8.1: AOPS conformance requirements to IEC 61511 according to IEC • AOPS reduces workload for operators
61511
AOPS is a safety instrumented function (SIF) and • IEC 61511 / 61508 offers equipment with
table 8.1 describes conformance to IEC 61511 as accreditation by third party assessors with
recommended for AOPS that are SIL 0, although standardized failure-rate data and safety
it is not required. The background is that safety manuals
requirements are continuously increasing, and what
is SIL 0 today may become SIL 1 in the future. Hence 8.4 Hardware fault tolerance
the future-proof approach is to design SIL 0 AOPS in An AOPS needs to consist of a sensor, a logic solver,
conformance with IEC 61511. and an actuator. However, it is a common practice
Similarly the upgrade of existing OPS is often a to add more than one of certain elements within
gradual process over several years where the sensors, the same AOPS. This is referred to as a system’s
logic-solver and actuators are upgraded in different Hardware Fault Tolerance (HFT) and can be employed
projects. The existing system may be a MOPS or an to increase both reliability and availability of an OPS.
AOPS that was designed before the first edition of IEC Hardware Fault Tolerance (HFT) can be employed
61511 was released in 2003. Often the requirements to both increase the reliability and availability of an
are uncertain. Maybe originally the goal is a risk OPS as described in the following examples. Figure
reduction factor of 10 to 100 (SIL 1) but later evolves 8.3 illustrates the most basic setup. A single sensor is
to 100 to 1,000 (SIL 2). The future-proof approach to connected to a single logic solver that communicates
the inherent uncertainty in many OPS projects is to with a single actuator. There are no redundant
select equipment from the beginning that: elements, hence HFT=0. This system is referred to
as 1oo1 (1-out-of-1) since each element single-
1. Can be used in AOPS conforming to IEC 61511 handedly determines the action of the system.
as described in section “Equipment selection”
chapter 5 An alternative approach is to add a second actuator
as illustrated in figure 8.4. There is 1 redundant
2. Can be used, or easily upgraded, to meet a actuator, which makes HFT=1 for this setup. It is
higher SIL than currently expected (target = SIL referred to as 1oo2 (1-out-of-2) since only 1 of the
requirement + 1) 2 actuators needs to successfully close in order to
Input to the selection of individual components prevent an overfill. This setup will increase reliability,
in an OPS can be found in chapter 10 “Equipment but decrease the availability.
selection”. A third, and increasingly common alternative is to
use a configuration of 2oo3 (2-out-of-3) sensors. The
8.3 AOPS vs. MOPS MOPS will close the valve when 2 of the 3 sensors
agree that it is the proper action to take. With 2
MOPS has traditionally been used in some redundant sensors, HFT increases to HFT=2, and in
applications because it is easier to implement, has comparison to a 1oo1 configuration, this provides
lower initial capital expenditure and less complexity. both increased reliability and availability.
However, modern overfill prevention takes
preference to AOPS in conformance with IEC 61511
rather than MOPS because:
• Humans are inherently unreliable, and
therefore MOPS is limited to a risk reduction
66
8 - Overfill prevention system
LT SIS
LT SIS
Figure 8.4: OPS consisting of 1oo2 actuators (HFT = 1). This configuration increases the reliability, but decreases the availability, compared to a 1oo1
configuration
LT
LT SIS
LT
Figure 8.5: OPS consisting of 2oo3 sensors (HFT = 2). This configuration increases both the reliability and availability, compared to a 1oo1 configuration
67
8 - Overfill prevention system
A critical aspect of overfill prevention is to correctly CH – LAHH = Max level rate x Response time +
define the levels of concern (LOC) which include Safety margin
Critical High (CH), Level Alarm High High (LAHH or
The location of the LAHH is commonly determined by
simply HiHi) and Maximum Working Level (MWL)
the following steps:
as depicted in figure 8.6 and described in table 8.2.
• The maximum level rate is calculated. Typically
based on the maximum flow-rate and the
diameter of the tank. Note that the diameter
in the tank may vary and this must be taken
into consideration
CH
• The response time is determined. This must
take the entire OPS into account. More
LAHH specifically:
MWL oo AOPS: the sum of the worst case response
times of the sensor, logic and actuator
oo MOPS: the sum of the worst case response
times of the level sensor, notification
system and subsequent manual actions.
The response time of the manual actions
may include the time for the operator to
observe the alarm, the time it takes to
communicate the alarm to a field operator,
time for a field operator to travel to the
actuator, and the time it takes to activate
Figure 8.6: The Levels Of Concern (LOC) for tank overfill prevention the actuator
• The safety margin to be used is defined, which
According to API 2350 the level alert high (LAH) is is ultimately a corporate decision
not included as a LOC but it may optionally be used
• Finally, LAHH is calculated by the following
for operational purposes. Note the difference in
formula: LAHH = CH - Max level rate x
terminology: LAHH is an alarm whereas LAH is an
Response time - Safety margin
alert. According to API 2350 an alarm is safety critical
and requires immediate action whereas alerts are Changes of the LOC should undergo a management
optional non-safety critical notifications. of change process, which is a part of the overfill
management system (OMS) described in chapter
Determining the LOC is a rigorous process where 7. Consequently, the LOC should not be changed
both internal and external requirements (chapter frequently or temporarily due to, for example,
5 “Industry standards” and chapter 4 “Regulatory operational inconveniences.
requirements”) should be taken into account as well
Table 8.2: API 2350 definition of The Levels Of Concern (LOC) for tank overfill prevention
68
9 - Proof-testing
9 Proof-testing
Topic
9.1
9.2
Proof-testing requirements
9.1.1
9.1.2
IEC 61511
API 2350
Proof-test interval
9.2.1
9.2.2
9.3
IEC 61511
API 2350
71
71
72
72
72
75
9.5 Implications 79
69
9 - Proof-testing
The purpose of proof-testing is to detect random • Ever-increasing need for safety and efficiency
hardware failures to verify that commissioned improvements
equipment already in operation functions correctly. • The introduction of IEC 61511 which
It is executed periodically and thereby differs from emphasizes the safety life-cycle approach
the site acceptance test (SAT) which is executed as a (figure 5.4) along with providing a theoretical
part of the commissioning or management of change framework for proof-testing and a quality
process to detect systematic (human) errors. metric (the coverage factor)
Proof-testing is a useful tool to reduce the probability • A number of high profile accidents where lack
of failure of infrequently used safety systems. It is of proper proof-testing was suspected to be
associated with the safety layer and not the BPCS one of the root-causes (e.g. the Buncefield
which is always in use and is therefore (at least accident)
theoretically) assumed to be continuously verified. The trend in the industry is to include proof-testing as
The BPCS may need periodic verification but this is a key selection criterion when purchasing equipment
typically not denoted proof-testing since the purpose since the cost to execute once the equipment has
is different (e.g. accuracy verification rather than been commissioned can be considerable. Other
detecting random hardware failures). In this guide, important aspects involve personnel and process
proof-testing is synonymous with verification of the safety.
overfill prevention system (OPS).
Proof-testing is generic and applies to any type
of equipment. It is critical that the entire safety
function and associated equipment are included. At a
minimum, there will be a sensor, actuator and a logic
solver, but for an OPS, this could be interpreted as
level sensors, a PLC, valves, emergency stop buttons,
and audiovisual alarms. See figure 9.1.
LT LT
SIS LC
Figure 9.1: Proof-testing applies to all components of an overfill prevention system (OPS)
70
9 - Proof-testing
71
9 - Proof-testing
API 2350 provides requirements for testing of overfill IEC 61511 provides a theoretical framework for the
prevention systems which are equally applicable to calculation of the proof-test interval. An important
both MOPS and AOPS. The requirements are similar fundamental assumption for that framework is that
to those found in IEC 61511, although targeted the random hardware failure rate of a level sensor
specifically towards the bulk liquid industry. The most is constant during its useful lifetime. This is often
important requirements are: referenced as the middle section of a so called
bathtub curve. The bathtub curve is a widely used
• Proof-test procedures shall be documented model in reliability engineering and a more detailed
and schedules for periodic proof-testing shall explanation is provided in figure 9.2.
be established Early Wear-out
• P
roof-test records shall be maintained for at Failures Random Failures Failures
Failure Rate
• T
he personnel executing the proof-testing
shall be competent. The facility is responsible
for assigning dedicated personnel and
providing appropriate training Time
Figure 9.2: The bathtub curve
9. 2 Proof-test interval
There are two basic methods for the determination of 9.2.1.2 Probability of failure on demand
a proof-test interval: According to IEC 61511, the proof-test interval shall
• P
rescriptive method with predetermined be calculated based on the average probability of
interval failure on demand, denoted PFDavg, during the time
that the safety function is in operation (mission time).
• A
nalytical method based on equipment For instance, an overfill prevention system with a high
reliability and required risk reduction PFDavg runs a high risk of failing to close a shutdown
The traditional approach is to use a predetermined valve in an event of excessive tank levels, whereas an
interval which may result in an over or under overfill prevention system with low PFDavg is more
engineered solution. The modern approach therefore reliable. The PFDavg value needs to match the required
uses the analytical method to calculate an interval risk reduction factor as described in table 9.1.
appropriate for the specific safety function.
SIL RRF PFDavg
In practice, a number of factors based on internal and
external requirements must be taken into account 1 10-100 0.1-0.01
when determining the proof-test interval. The 2 100-1,000 0.01-0.001
remainder of this section describes the requirements
according to IEC 61511 and API 2350. 3 1,000-10,000 0.001-0.0001
4 10,000-100,000 0.0001-0.00001
9.2.1 IEC 61511
Table 9.1: Risk reduction factors (RRF) and average probability of failure on
demand (PFDavg) segmented by safety integrity levels(SIL)
According to the IEC 61511 methodology, the most
important factors affecting the proof-test interval Calculating PFDavg involves a multitude of factors.
are: Software packages exist with complex models
• The safety functions risk reduction factor (RRF) but IEC 61508-6 provides approximate simplified
formulas. Assuming non-redundant configurations
• The reliability of the device (λDU) (1oo1) where λDU is the safety function’s dangerous
• P
roof-test effectiveness (coverage factor) and undetected failure rate and T is the time interval:
existence of partial proof-testing
PFD ≈ λDU* T
• M
ission time, i.e. the time from a system’s
start-up until its replacement or refurbishment PFDavg ≈ λDU* T / 2
to as-new condition
The risk reduction factor (RRF) can be calculated in
the following way:
RRF = 1/PFDavg
72
9 - Proof-testing
Time
73
9 - Proof-testing
PFD(avg)=0.15% Measurement
electronics
T
FI
80
Example 9.2: Estimating the proof-test interval for a safety function’s Process Process
sensor
74
9 - Proof-testing
Usually, partial proof-tests are used to extend the often non-documented procedures have been
time interval of the comprehensive proof-test. cumbersome and in some cases dangerous and
Mathematically, the partial proof-test has a lower resulted in considerable downtime. Documented
coverage factor than the comprehensive proof-test. evidence that the proof-test has been executed
The principal effect on the probability of failure on correctly is often incomplete or nonexisting.
demand is depicted in figure 9.9.
9.3.1 Traditional proof-testing procedures
Probability of exemplified with point level sensors
failure
Comprehensive
Proof-test Although the trend is towards using continuous level
sensors for safety critical measurements, point-
Partial Proof-test
level sensors have been traditionally used for these
types of applications. Over the years, equipment
manufacturers, system integrators and users have
developed several different proof-testing procedures,
which can broadly be separated into the categories
listed below and overleaf.
Time
Figure 9.9: Test coverage of partial and comprehensive proof-testing
75
9 - Proof-testing
Bucket testing
Another traditional proof-testing method is to dismount the point level sensor and expose it to the alarm
condition. In practice, this is often performed by inserting the device into a bucket filled with product. This
method requires a visit to the tank and access to the level sensor while the tank is temporarily taken out of
operation. The procedure may be a direct safety concern to the personnel executing the test since it both
exposes the tank to the atmosphere and the bucket contents may be hazardous. Additional precautions
must be taken if it is a pressurized tank or an explosive environment. Ideally, the product in the bucket
should be the same as in the tank, but for safety reasons, water is often used.
When the test is not performed with the media to be
measured, there is an obvious risk that test results
become irrelevant for the true process conditions.
Furthermore, when sensors are dismounted, there
is no guarantee that re-commissioning is correctly
executed. There may be cable glitches, gaskets
missing, loose bolts or even damage imposed to the
sensor itself.
One advantage with this type of testing however,
is that it allows for visual inspection of the sensor’s
wetted parts. For example indications of corrosion
or material incompatibility may be used as input for
predictive maintenance.
76
9 - Proof-testing
77
9 - Proof-testing
Table 9.2: Description of selected radar level sensors’ proof-test procedure segmented by its major components
Figure 9.10: Example of better proof-testing methods with modern overfill prevention
equipment
78
9 - Proof-testing
9.5 Implications
Proof-testing has become an increasingly important
feature and is now one of the key selection criteria
when selecting equipment for modern overfill
prevention systems. Some of the relevant features
are:
• I s the proof-test procedure properly
described?
• A
re both comprehensive and partial proof-
tests available?
• H
as the proof-test been assessed by an
accredited 3rd party?
• Is the proof-test IEC 61508 certified?
• Quantitative justification:
oo I s the effectiveness (coverage factor)
specified?
oo Is the failure-rate (λ) specified?
79
9 - Proof-testing
80
10 - Equipment selection
10 Equipment Selection
Topic
10.1
Overfill prevention system
10.1.1 Level sensors
Page
82
82
81
10 - Equipment selection
82
10 - Equipment selection
83
10 - Equipment selection
84
11 - Appendix: Overfill prevention systems examples
11 Appendix: Overfill
prevention system
examples
Topic
11.1
Bulk liquid storage
11.1. 1 Fixed roof tanks
11.1. 2 Floating roof tanks
11.1. 3 Spherical tanks
11.1. 4 Bullet tanks
Page
86
86
88
90
92
94
11.2. 1 Top mounted - OPS level sensor 94
11.2. 2 Chamber mounted - OPS level sensor 97
11.2. 3 Side mounted - OPS level sensor 99
11.2. 4 Separator tank 101
11.2. 5 Distillation column 102
11.2. 6 Boiler drum 103
11.2. 7 Reactor tank 104
85
11 - Appendix: Overfill prevention systems examples
Rosemount
Rosemount 2240S with Rosemount
5900S Radar Multiple Point 5900S Radar
Level Gauge Temperature Level Gauge
Display
SI
High-High Alarm
Delta V
SIS
Smart
3
Rosemount 2410
L
SI
Wireless
Tank Hub
THUM™
Adapter
SIL-PAC
(Fisher DVC + Fisher Rosemount 2410
Bettis Actuator) Valve Tank Hub
SI
L3
1400 Smart
TankMaster Inventory Wireless Gateway
Management Software
Connection to
TankMaster Rosemount 2460
(optional) System Hub
Illustration shows a fixed roof tank equipped with Automatic Tank Gauging based on the Rosemount 5900S
and a SIL 3 AOPS based on the Rosemount 5900S, DeltaV SIS and a Bettis actuator.
86
11 - Appendix: Overfill prevention systems examples
Below are alternative Rosemount level sensors for fixed roof tanks:
2
SI
L L2
SI
L2
SI
L2
SI
Level
2
SIL
L2
SI
L3
SI
Overfill
Level
Rosemount
2
SIL 2240S with
Multiple Point
Rosemount Rosemount 2230
Temperature
Independent 5900S 2-in-1 Graphical Field
Alarm Panel Radar Level Display
Gauge
High-High Alarm
Delta V
SIS
L3
Rosemount 2410
SI
1400 Smart
TankMaster Inventory Wireless Gateway
Management Software
Connection to
TankMaster Rosemount 2460
(optional) System Hub
Illustration shows a floating roof tank equipped with Automatic Tank Gauging based on the Rosemount 5900S
and a SIL 3 AOPS based on the Rosemount 5900S, DeltaV SIS and a Bettis actuator.
88
11 - Appendix: Overfill prevention systems examples
Below are alternative Rosemount level sensors for floating roof tanks:
L2
SI
L2
SI
L3
SIL
SI
2
Level
2
SIL
SIL
2
SI
L
3
89
11 - Appendix: Overfill prevention systems examples
Rosemount Rosemount
5900S Radar 5900S Radar
Level Gauge Level Gauge
Independent with Pressure with Pressure
Transmitter Transmitter Rosemount 2230
Alarm Panel Graphical Field
L3
SI Display
Rosemount 644
with Single Point
High-High Alarm
Temperature
Delta V
Verification
SIS
Pin
Rosemount 644
L3
1400 Smart
TankMaster Inventory Wireless Gateway
Management Software
Connection to
TankMaster Rosemount 2460
(optional) System Hub
Illustration shows a spherical tank equipped with Automatic Tank Gauging based on the Rosemount 5900S
and a SIL 3 AOPS based on the Rosemount 5900S, DeltaV SIS and a Bettis actuator.
90
11 - Appendix: Overfill prevention systems examples
2
SIL
Overfill
Level
2
SIL
Rosemount
5900S 2-in-1 (AOPS, MOPS)
91
11 - Appendix: Overfill prevention systems examples
Independent
L3
Rosemount Rosemount 2230
SI
Alarm Panel Graphical Field
5900S Radar
Rosemount Display
5900S Radar Level Gauge
Level Gauge with Pressure
with Pressure Transmitter
High-High Alarm
Transmitter
Delta V
SIS Rosemount
644 with
Single Point
Temperature
L3
Rosemount 2410
SI
Tank Hub
Verification
SIL-PAC Pin Rosemount 2410
(Fisher DVC + Fisher
Tank Hub
Bettis Actuator) Valve
Smart Wireless
THUM™ Adapter
SI
L3
1400 Smart
TankMaster Inventory Wireless Gateway
Management Software
Connection to
TankMaster Rosemount 2460
(optional) System Hub
Illustration shows a bullet tank equipped with Automatic Tank Gauging based on the Rosemount 5900S and a
SIL 3 AOPS based on the Rosemount 5900S, DeltaV SIS and a Bettis actuator.
92
11 - Appendix: Overfill prevention systems examples
2
SIL
2
SI
L L2
SI
L3
SI
Level
2
SIL
Rosemount
5900S 2-in-1 (AOPS, MOPS)
Additional bulk liquid storage tank examples is available in “The Complete Guide to API 2350”
(Ref.No. 901030)
93
11 - Appendix: Overfill prevention systems examples
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
SIL 2
Control/
Safety Logic Solver Monitoring System
HiHi
SIL 3 Hi
MWL
3
SIL
ESD-Valve Pump
MaxWL - Maximum Work Level Hi - High Level HiHi - High High Level
Illustration shows a cone tank equipped with a Rosemount 5400 for BPCS and SIL 2 AOPS based on Rosemount
5300, DeltaV SIS and Bettis actuator.
94
11 - Appendix: Overfill prevention systems examples
Manual Overfill Prevention System (MOPS) Basic Process Control System (BPCS)
Control/
Monitoring System
HiHi
Hi
Audible Alarm Visual Alarm
MWL
MWL - Maximum Work Level Hi - High Level HiHi - High High Level
Illustration shows a cone tank equipped with a Rosemount 5300 for BPCS and MOPS based on a Rosemount
2100.
95
11 - Appendix: Overfill prevention systems examples
SIL 2
SIL 2
96
11 - Appendix: Overfill prevention systems examples
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
SIL 3
SIL 2
Control/
HiHi
Monitoring System
Hi
MWL
ESD-Valve Pump
MWL - Maximum Work Level Hi - High Level HiHi - High High Level
Illustration shows chamber installations. Rosemount 5300 is used for BPCS and SIL 2 AOPS are based on
Rosemount 5300, DeltaV SIS and Bettis actuator.
Rosemount Rosemount
2100 column (MOPS) 5300
97
11 - Appendix: Overfill prevention systems examples
Manual Overfill Prevention System (MOPS) Basic Process Control System (BPCS)
Control/
HiHi
Monitoring System
Hi
Audible Alarm Visual Alarm
MWL
MWL - Maximum Work Level Hi - High Level HiHi - High High Level
Illustration shows a cone tank equipped with a Rosemount 5300 for BPCS and MOPS based on a
Rosemount 3308.
Rosemount Rosemount
2160 column (MOPS) 5300
98
11 - Appendix: Overfill prevention systems examples
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
Control/
Monitoring System
HiHi
Hi
SIL 2
SIL 3
3
SIL
ESD-Valve Pump
MWL - Maximum Work Level Hi - High Level HiHi - High High Level
Illustration shows a tank side installation. Rosemount 5300 is used for BPCS and SIL 2 AOPS is based on
Rosemount 2100, DeltaV SIS and Bettis actuator.
99
11 - Appendix: Overfill prevention systems examples
Manual Overfill Prevention System (MOPS) Basic Process Control System (BPCS)
Control/
Monitoring System
HiHi
Hi
MWL
MWL - Maximum Work Level Hi - High Level HiHi - High High Level
Illustration shows a tank side installation. Rosemount 5300 for BPCS and MOPS based on a wireless Rosemount
vertical float switch 702.
Below is an alternative Rosemount level sensor for side mounting:
Rosemount Rosemount
2160 (MOPS) 5300
100
11 - Appendix: Overfill prevention systems examples
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
Safety Logic Solver Control/
Monitoring System
SIL 3
Audible Visual
Alarm Alarm
3
SIL
ESD-Valve Pump
SIL
2
SIL
2
Illustration shows a separator tank equipped BPCS with two Rosemount 5300 for level and interface
measurement and SIL2 AOPS and SIL2 dry-run protection based on Rosemount 2100, DeltaV SIS and
Bettis actuator.
101
11 - Appendix: Overfill prevention systems examples
11.2.4 Distillation Column rise through the column, different components will
condense at different temperatures and accumulate
Distillation columns allow separation of fluid for withdrawal.
mixtures based upon their boiling points. As vapors
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
Control/
Monitoring System
SIL 3
SIL 2
SIL 2
SIL 2
3
SIL SIL 2
ESD-Valve Pump
SIL 2
Illustration shows a distillation column equipped with a BPCS with a Rosemount 5300 for level measurement
and SIL2 AOPS based on Rosemount 5300, DeltaV SIS and Bettis actuator.
102
11 - Appendix: Overfill prevention systems examples
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
Safety Logic Solver
Control/
SIL 3 Monitoring System
Audible Visual
Alarm Alarm
3
SIL
ESD-Valve Pump
Illustration shows a boiler drum equipped with a BPCS with a Rosemount 5300 for level measurement and SIL3
AOPS based on three Rosemount 5300 (2oo3), DeltaV SIS and Bettis actuator.
103
11 - Appendix: Overfill prevention systems examples
Automatic Overfill Prevention System (AOPS) Basic Process Control System (BPCS)
Control/
Safety Logic Solver
Monitoring System
SIL 3
SIL 2
Audible Visual
Alarm Alarm
3
SIL
ESD-Valve Pump
Illustration shows a blending tank equipped with a BPCS of Rosemount differential pressure level measurement
gauge and SIL2 AOPS based on the Rosemount 5400, DeltaV SIS and Bettis actuctor.
104
12 - References
12 References
Topic
12.1
12.2
Literature references
Picture references
Page
106
106
105
12 - References
12. References
12.1 Literature references
American Petroleum Institute (2012) API 2350. Overfill Protection for Storage Tanks in Petroleum Facilities,
Fourth edition
Central Intelligence Agency (2015) The World Factbook, https://www.cia.gov/library/publications/the-world-
factbook/rankorder/2246rank.html 2015-09-04
Center for Chemical Process Safety (2007) Guidelines for Risk Based Process Safety, Wiley
Control of Major Accident Hazards (2011) Buncefield: Why did it happened? http://www.hse.gov.uk/comah/
buncefield/buncefield-report.pdf 2015-09-03
Felten, D., (2015) When Prevention Fails: Managing Your Spill Response, NISTM 17th Annual International
Aboveground Storage Tank Conference & Trade Show, Florida
Goble, W., (2013) Make the IEC 61511 into a cookbook? http://www.exida.com/Blog/Make-IEC-61511-into-a-
Cookbook#sthash.oqiYamB1.dpuf 2015-07-21
International Electrotechnical Commission () IEC 61511 Functional safety - Safety instrumented systems for
the process industry sector
International Electrotechnical Commission (2010) IEC 61508 Functional Safety of Electrical/Electronic/
Programmable Electronic Safety-related Systems
Loren (2005) BP reveals costs of Texas City settlements, http://blog.chron.com/lorensteffy/2005/07/bp-
reveals-costs-of-texas-city-settlements/ 15-07-17
Marsh & McLennan Companies (2011) Management of Atmospheric Storage Tank, Rev 01, United Kingdom
Mars (2007) Recommendations on the design and operation of fuel storage sites, Buncefield Major Incident
Investigation Board
M B Lal Committee Report (2009) Independent Inquiry Committee Report on Indian Oil Terminal Fire at Jaipur
http://oisd.gov.in/ 2015-09-04
Puerto Rico Seismic Network (2009) Informe Especial, Explosión de Caribbean Petroleum en Bayamón,
University of Puerto Rico Mayagüez Campus.
Sreenevasan, R., (2015) The effect of regulations in improving process safety, Tetra Tech Proteus, https://
www.engineersaustralia.org.au/sites/default/files/shado/Learned%20Groups/Technical%20Societies/Risk%20
Engineering%20Society/australian_regulations_res_wa_paper.pdf 2015-07-21
United States Environmental Protection Agency (2014) Response to Oil Spills, http://www.epa.gov/ceppo/
web/content/learning/response.htm 2015-07-14
U.S. Chemical Safety and Hazard Investigation Board (2015) FINAL INVESTIGATION REPORT
CARIBBEAN PETROLEUM TANK TERMINAL EXPLOSION AND MULTIPLE TANK FIRES http://www.csb.gov/
assets/1/16/06.09.2015_FINAL_CAPECO_Draft_Report__for_Board_Vote.pdf 2015-09-03
106
12 - References
107
12 - References
108
About the authors
Lead author
Carl-Johan Roos, Functional Safety Officer, Emerson Process
Management, Rosemount Level
Carl-Johan “CJ” Roos has 10+ years of global experience from various
technical and managerial positions in the process industry. Besides
API2350, he has actively participated in numerous product specific
IEC61508 certifications and site specific IEC61511 related projects,
besides his usual work as functional safety officer at Emerson’s Process
Level division where he often addresses national overfill prevention
regulations such as TÜV/DIBt WHG in Germany. In 2015, he was awarded
with Process Control Magazine’s award “Engineering Leaders Under
40”. Roos has a Master’s degree in Electrical and Computer Engineering
from Georgia Tech and Chalmers University, and a Master’s of Business
Administration degree from the University of Gothenburg.
Co-author
Phil E. Myers, API 2350 Committee Chairman
Phil E. Myers has chaired numerous task groups for the American
Petroleum Institute, including API 2350. Currently, he is the director of
PEMY Consulting. He has also worked at Chevron Corporation where
he was a mechanical integrity specialist for tanks, piping and pressure
vessels specializing in safety and risk. Myers holds a BSc in Chemical
Engineering from UC Berkeley and an MSc in Theoretical and Applied
Statistics from California State University.
Acknowledgements
This handbook is the result of a joint effort of Emerson colleagues and customers around the world.
Thanks to all of the Emerson functional safety experts who gave their input to this project, and laid the
foundation of the content.
Thank you also to Patricia Mattsson and Martin Larsson for developing the outstanding visuals and layout to
enhance the user experience of this handbook. A special thanks also goes out to Peta Glenister for her editing
and valuable comments.
Finally, thanks to all of the unnamed contributors and all of the Rosemount Level users out there!
Rosemount products for overfill prevention
Process Industry
Automatic Overfill Prevention System Basic Process Control Manual Overfill Prevention System
System
2100
5300
5400 5400
5300 Wireless
Mobrey Float
2100 Switches
SIL 3
Control/
Monitoring System
Audible
Alarm
Visual
Alarm
3
SIL
ESD-Valve
Pump SIL 2 SIL 2
SIL 3
Control/
Monitoring System
Audible
Alarm
Visual
Alarm
3
SIL
ESD-Valve
Pump SIL 2 SIL 2
Introduction
Why invest?
Key elements
Regulatory requirements
Industry standards
Risk assessment
Proof-testing
Equipment selection
References
Global capabilities
Global Headquarters Europe Regional Office
Emerson Process Management Emerson Process Management Europe GmbH
6021 Innovation Blvd. Neuhofstrasse 19a P.O. Box 1046
Shakopee, MN 55379, USA CH 6340 Baar
+1 800 999 9307 or +1 952 906 8888 Switzerland
+1 952 949 7001 +41 (0) 41 768 6111
RFQ.RMD-RCC@EmersonProcess.com +41 (0) 41 768 6300
RFQ.RMD-RCC@EmersonProcess.com
Asia Pacific Regional Office Middle East and Africa Regional Office
Emerson Process Management Asia Pacific Pte Ltd Emerson Process Management
1 Pandan Crescent Emerson FZE P.O. Box 17033,
Singapore 128461 Jebel Ali Free Zone - South 2
+65 6777 8211 Dubai, United Arab Emirates
+65 6777 0947 +971 4 8118100
Enquiries@AP.EmersonProcess.com +971 4 8865465
RFQ.RMTMEA@Emerson.com
www.rosemount.com
ISBN 9789198277906
57599 >
9 789198 277906